April 1, 2020

3452 words 17 mins read

Paper Group ANR 472

Semantic Adversarial Perturbations using Learnt Representations. On Constraint Definability in Tractable Probabilistic Models. Pipeline Interventions. An Analysis of Adversarial Attacks and Defenses on Autonomous Driving Models. Stochastic Runge-Kutta methods and adaptive SGD-G2 stochastic gradient descent. Kernel and Rich Regimes in Overparametriz …

Semantic Adversarial Perturbations using Learnt Representations


Title	Semantic Adversarial Perturbations using Learnt Representations
Authors	Isaac Dunn, Tom Melham, Daniel Kroening
Abstract	Adversarial examples for image classifiers are typically created by searching for a suitable norm-constrained perturbation to the pixels of an image. However, such perturbations represent only a small and rather contrived subset of possible adversarial inputs; robustness to norm-constrained pixel perturbations alone is insufficient. We introduce a novel method for the construction of a rich new class of semantic adversarial examples. Leveraging the hierarchical feature representations learnt by generative models, our procedure makes adversarial but realistic changes at different levels of semantic granularity. Unlike prior work, this is not an ad-hoc algorithm targeting a fixed category of semantic property. For instance, our approach perturbs the pose, location, size, shape, colour and texture of the objects in an image without manual encoding of these concepts. We demonstrate this new attack by creating semantic adversarial examples that fool state-of-the-art classifiers on the MNIST and ImageNet datasets.
Tasks
Published	2020-01-29
URL	https://arxiv.org/abs/2001.11055v1
PDF	https://arxiv.org/pdf/2001.11055v1.pdf
PWC	https://paperswithcode.com/paper/semantic-adversarial-perturbations-using
Repo
Framework

On Constraint Definability in Tractable Probabilistic Models


Title	On Constraint Definability in Tractable Probabilistic Models
Authors	Ioannis Papantonis, Vaishak Belle
Abstract	Incorporating constraints is a major concern in probabilistic machine learning. A wide variety of problems require predictions to be integrated with reasoning about constraints, from modelling routes on maps to approving loan predictions. In the former, we may require the prediction model to respect the presence of physical paths between the nodes on the map, and in the latter, we may require that the prediction model respect fairness constraints that ensure that outcomes are not subject to bias. Broadly speaking, constraints may be probabilistic, logical or causal, but the overarching challenge is to determine if and how a model can be learnt that handles all the declared constraints. To the best of our knowledge, this is largely an open problem. In this paper, we consider a mathematical inquiry on how the learning of tractable probabilistic models, such as sum-product networks, is possible while incorporating constraints.
Tasks
Published	2020-01-29
URL	https://arxiv.org/abs/2001.11349v1
PDF	https://arxiv.org/pdf/2001.11349v1.pdf
PWC	https://paperswithcode.com/paper/on-constraint-definability-in-tractable
Repo
Framework

Pipeline Interventions


Title	Pipeline Interventions
Authors	Eshwar Ram Arunachaleswaran, Sampath Kannan, Aaron Roth, Juba Ziani
Abstract	We introduce the \emph{pipeline intervention} problem, defined by a layered directed acyclic graph and a set of stochastic matrices governing transitions between successive layers. The graph is a stylized model for how people from different populations are presented opportunities, eventually leading to some reward. In our model, individuals are born into an initial position (i.e. some node in the first layer of the graph) according to a fixed probability distribution, and then stochastically progress through the graph according to the transition matrices, until they reach a node in the final layer of the graph; each node in the final layer has a \emph{reward} associated with it. The pipeline intervention problem asks how to best make costly changes to the transition matrices governing people’s stochastic transitions through the graph, subject to a budget constraint. We consider two objectives: social welfare maximization, and a fairness-motivated maximin objective that seeks to maximize the value to the population (starting node) with the \emph{least} expected value. We consider two variants of the maximin objective that turn out to be distinct, depending on whether we demand a deterministic solution or allow randomization. For each objective, we give an efficient approximation algorithm (an additive FPTAS) for constant width networks. We also tightly characterize the “price of fairness” in our setting: the ratio between the highest achievable social welfare and the highest social welfare consistent with a maximin optimal solution. Finally we show that for polynomial width networks, even approximating the maximin objective to any constant factor is NP hard, even for networks with constant depth. This shows that the restriction on the width in our positive results is essential.
Tasks
Published	2020-02-16
URL	https://arxiv.org/abs/2002.06592v1
PDF	https://arxiv.org/pdf/2002.06592v1.pdf
PWC	https://paperswithcode.com/paper/pipeline-interventions
Repo
Framework

An Analysis of Adversarial Attacks and Defenses on Autonomous Driving Models


Title	An Analysis of Adversarial Attacks and Defenses on Autonomous Driving Models
Authors	Yao Deng, Xi Zheng, Tianyi Zhang, Chen Chen, Guannan Lou, Miryung Kim
Abstract	Nowadays, autonomous driving has attracted much attention from both industry and academia. Convolutional neural network (CNN) is a key component in autonomous driving, which is also increasingly adopted in pervasive computing such as smartphones, wearable devices, and IoT networks. Prior work shows CNN-based classification models are vulnerable to adversarial attacks. However, it is uncertain to what extent regression models such as driving models are vulnerable to adversarial attacks, the effectiveness of existing defense techniques, and the defense implications for system and middleware builders. This paper presents an in-depth analysis of five adversarial attacks and four defense methods on three driving models. Experiments show that, similar to classification models, these models are still highly vulnerable to adversarial attacks. This poses a big security threat to autonomous driving and thus should be taken into account in practice. While these defense methods can effectively defend against different attacks, none of them are able to provide adequate protection against all five attacks. We derive several implications for system and middleware builders: (1) when adding a defense component against adversarial attacks, it is important to deploy multiple defense methods in tandem to achieve a good coverage of various attacks, (2) a blackbox attack is much less effective compared with a white-box attack, implying that it is important to keep model details (e.g., model architecture, hyperparameters) confidential via model obfuscation, and (3) driving models with a complex architecture are preferred if computing resources permit as they are more resilient to adversarial attacks than simple models.
Tasks	Autonomous Driving
Published	2020-02-06
URL	https://arxiv.org/abs/2002.02175v1
PDF	https://arxiv.org/pdf/2002.02175v1.pdf
PWC	https://paperswithcode.com/paper/an-analysis-of-adversarial-attacks-and
Repo
Framework

Stochastic Runge-Kutta methods and adaptive SGD-G2 stochastic gradient descent


Title	Stochastic Runge-Kutta methods and adaptive SGD-G2 stochastic gradient descent
Authors	Imen Ayadi, Gabriel Turinici
Abstract	The minimization of the loss function is of paramount importance in deep neural networks. On the other hand, many popular optimization algorithms have been shown to correspond to some evolution equation of gradient flow type. Inspired by the numerical schemes used for general evolution equations we introduce a second order stochastic Runge Kutta method and show that it yields a consistent procedure for the minimization of the loss function. In addition it can be coupled, in an adaptive framework, with a Stochastic Gradient Descent (SGD) to adjust automatically the learning rate of the SGD, without the need of any additional information on the Hessian of the loss functional. The adaptive SGD, called SGD-G2, is successfully tested on standard datasets.
Tasks
Published	2020-02-20
URL	https://arxiv.org/abs/2002.09304v1
PDF	https://arxiv.org/pdf/2002.09304v1.pdf
PWC	https://paperswithcode.com/paper/stochastic-runge-kutta-methods-and-adaptive
Repo
Framework

Kernel and Rich Regimes in Overparametrized Models


Title	Kernel and Rich Regimes in Overparametrized Models
Authors	Blake Woodworth, Suriya Gunasekar, Jason D. Lee, Edward Moroshko, Pedro Savarese, Itay Golan, Daniel Soudry, Nathan Srebro
Abstract	A recent line of work studies overparametrized neural networks in the “kernel regime,” i.e. when the network behaves during training as a kernelized linear predictor, and thus training with gradient descent has the effect of finding the minimum RKHS norm solution. This stands in contrast to other studies which demonstrate how gradient descent on overparametrized multilayer networks can induce rich implicit biases that are not RKHS norms. Building on an observation by Chizat and Bach, we show how the scale of the initialization controls the transition between the “kernel” (aka lazy) and “rich” (aka active) regimes and affects generalization properties in multilayer homogeneous models. We also highlight an interesting role for the width of a model in the case that the predictor is not identically zero at initialization. We provide a complete and detailed analysis for a family of simple depth-$D$ models that already exhibit an interesting and meaningful transition between the kernel and rich regimes, and we also demonstrate this transition empirically for more complex matrix factorization models and multilayer non-linear networks.
Tasks
Published	2020-02-20
URL	https://arxiv.org/abs/2002.09277v2
PDF	https://arxiv.org/pdf/2002.09277v2.pdf
PWC	https://paperswithcode.com/paper/kernel-and-rich-regimes-in-overparametrized-1
Repo
Framework

Neural Style Difference Transfer and Its Application to Font Generation


Title	Neural Style Difference Transfer and Its Application to Font Generation
Authors	Gantugs Atarsaikhan, Brian Kenji Iwana, Seiichi Uchida
Abstract	Designing fonts requires a great deal of time and effort. It requires professional skills, such as sketching, vectorizing, and image editing. Additionally, each letter has to be designed individually. In this paper, we will introduce a method to create fonts automatically. In our proposed method, the difference of font styles between two different fonts is found and transferred to another font using neural style transfer. Neural style transfer is a method of stylizing the contents of an image with the styles of another image. We proposed a novel neural style difference and content difference loss for the neural style transfer. With these losses, new fonts can be generated by adding or removing font styles from a font. We provided experimental results with various combinations of input fonts and discussed limitations and future development for the proposed method.
Tasks	Style Transfer
Published	2020-01-21
URL	https://arxiv.org/abs/2001.07321v1
PDF	https://arxiv.org/pdf/2001.07321v1.pdf
PWC	https://paperswithcode.com/paper/neural-style-difference-transfer-and-its
Repo
Framework

Towards Better Surgical Instrument Segmentation in Endoscopic Vision: Multi-Angle Feature Aggregation and Contour Supervision


Title	Towards Better Surgical Instrument Segmentation in Endoscopic Vision: Multi-Angle Feature Aggregation and Contour Supervision
Authors	Fangbo Qin, Shan Lin, Yangming Li, Randall A. Bly, Kris S. Moe, Blake Hannaford
Abstract	Accurate and real-time surgical instrument segmentation is important in the endoscopic vision of robot-assisted surgery, and significant challenges are posed by frequent instrument-tissue contacts and continuous change of observation perspective. For these challenging tasks more and more deep neural networks (DNN) models are designed in recent years. We are motivated to propose a general embeddable approach to improve these current DNN segmentation models without increasing the model parameter number. Firstly, observing the limited rotation-invariance performance of DNN, we proposed the Multi-Angle Feature Aggregation (MAFA) method, lever-aging active image rotation to gain richer visual cues and make the prediction more robust to instrument orientation changes. Secondly, in the end-to-end training stage, the auxiliary contour supervision is utilized to guide the model to learn the boundary awareness, so that the contour shape of segmentation mask is more precise. The effectiveness of the proposed methods is validated with ablation experiments con-ducted on novel Sinus-Surgery datasets.
Tasks
Published	2020-02-25
URL	https://arxiv.org/abs/2002.10675v1
PDF	https://arxiv.org/pdf/2002.10675v1.pdf
PWC	https://paperswithcode.com/paper/towards-better-surgical-instrument
Repo
Framework

Universal Semantic Segmentation for Fisheye Urban Driving Images


Title	Universal Semantic Segmentation for Fisheye Urban Driving Images
Authors	Yaozu Ye, Kailun Yang, Kaite Xiang, Juan Wang, Kaiwei Wang
Abstract	Semantic segmentation is a critical method in the field of autonomous driving. When performing semantic image segmentation, a wider field of view (FoV) helps to obtain more information about the surrounding environment, making automatic driving safer and more reliable, which could be offered by fisheye cameras. However, large public fisheye data sets are not available, and the fisheye images captured by the fisheye camera with large FoV comes with large distortion, so commonly-used semantic segmentation model cannot be directly utilized. In this paper, a seven degrees of freedom (DoF) augmentation method is proposed to transform rectilinear image to fisheye image in a more comprehensive way. In the training process, rectilinear images are transformed into fisheye images in seven DoF, which simulates the fisheye images taken by cameras of different positions, orientations and focal lengths. The result shows that training with the seven-DoF augmentation can evidently improve the model’s accuracy and robustness against different distorted fisheye data. This seven-DoF augmentation provides an universal semantic segmentation solution for fisheye cameras in different autonomous driving applications. Also, we provide specific parameter settings of the augmentation for autonomous driving. At last, we tested our universal semantic segmentation model on real fisheye images and obtained satisfactory results. The code and configurations are released at \url{https://github.com/Yaozhuwa/FisheyeSeg}.
Tasks	Autonomous Driving, Semantic Segmentation
Published	2020-01-31
URL	https://arxiv.org/abs/2002.03736v1
PDF	https://arxiv.org/pdf/2002.03736v1.pdf
PWC	https://paperswithcode.com/paper/universal-semantic-segmentation-for-fisheye
Repo
Framework

SummaryNet: A Multi-Stage Deep Learning Model for Automatic Video Summarisation


Title	SummaryNet: A Multi-Stage Deep Learning Model for Automatic Video Summarisation
Authors	Ziyad Jappie, David Torpey, Turgay Celik
Abstract	Video summarisation can be posed as the task of extracting important parts of a video in order to create an informative summary of what occurred in the video. In this paper we introduce SummaryNet as a supervised learning framework for automated video summarisation. SummaryNet employs a two-stream convolutional network to learn spatial (appearance) and temporal (motion) representations. It utilizes an encoder-decoder model to extract the most salient features from the learned video representations. Lastly, it uses a sigmoid regression network with bidirectional long short-term memory cells to predict the probability of a frame being a summary frame. Experimental results on benchmark datasets show that the proposed method achieves comparable or significantly better results than the state-of-the-art video summarisation methods.
Tasks
Published	2020-02-19
URL	https://arxiv.org/abs/2002.09424v1
PDF	https://arxiv.org/pdf/2002.09424v1.pdf
PWC	https://paperswithcode.com/paper/summarynet-a-multi-stage-deep-learning-model
Repo
Framework

Survey of Deep Reinforcement Learning for Motion Planning of Autonomous Vehicles


Title	Survey of Deep Reinforcement Learning for Motion Planning of Autonomous Vehicles
Authors	Szilárd Aradi
Abstract	Academic research in the field of autonomous vehicles has reached high popularity in recent years related to several topics as sensor technologies, V2X communications, safety, security, decision making, control, and even legal and standardization rules. Besides classic control design approaches, Artificial Intelligence and Machine Learning methods are present in almost all of these fields. Another part of research focuses on different layers of Motion Planning, such as strategic decisions, trajectory planning, and control. A wide range of techniques in Machine Learning itself have been developed, and this article describes one of these fields, Deep Reinforcement Learning (DRL). The paper provides insight into the hierarchical motion planning problem and describes the basics of DRL. The main elements of designing such a system are the modeling of the environment, the modeling abstractions, the description of the state and the perception models, the appropriate rewarding, and the realization of the underlying neural network. The paper describes vehicle models, simulation possibilities and computational requirements. Strategic decisions on different layers and the observation models, e.g., continuous and discrete state representations, grid-based, and camera-based solutions are presented. The paper surveys the state-of-art solutions systematized by the different tasks and levels of autonomous driving, such as car-following, lane-keeping, trajectory following, merging, or driving in dense traffic. Finally, open questions and future challenges are discussed.
Tasks	Autonomous Driving, Autonomous Vehicles, Decision Making, Motion Planning
Published	2020-01-30
URL	https://arxiv.org/abs/2001.11231v1
PDF	https://arxiv.org/pdf/2001.11231v1.pdf
PWC	https://paperswithcode.com/paper/survey-of-deep-reinforcement-learning-for
Repo
Framework

Segmenting Transparent Objects in the Wild


Title	Segmenting Transparent Objects in the Wild
Authors	Enze Xie, Wenjia Wang, Wenhai Wang, Mingyu Ding, Chunhua Shen, Ping Luo
Abstract	Transparent objects such as windows and bottles made by glass widely exist in the real world. Segmenting transparent objects is challenging because these objects have diverse appearance inherited from the image background, making them had similar appearance with their surroundings. Besides the technical difficulty of this task, only a few previous datasets were specially designed and collected to explore this task and most of the existing datasets have major drawbacks. They either possess limited sample size such as merely a thousand of images without manual annotations, or they generate all images by using computer graphics method (i.e. not real image). To address this important problem, this work proposes a large-scale dataset for transparent object segmentation, named Trans10K, consisting of 10,428 images of real scenarios with carefully manual annotations, which are 10 times larger than the existing datasets. The transparent objects in Trans10K are extremely challenging due to high diversity in scale, viewpoint and occlusion as shown in Fig. 1. To evaluate the effectiveness of Trans10K, we propose a novel boundary-aware segmentation method, termed TransLab, which exploits boundary as the clue to improve segmentation of transparent objects. Extensive experiments and ablation studies demonstrate the effectiveness of Trans10K and validate the practicality of learning object boundary in TransLab. For example, TransLab significantly outperforms 20 recent object segmentation methods based on deep learning, showing that this task is largely unsolved. We believe that both Trans10K and TransLab have important contributions to both the academia and industry, facilitating future researches and applications.
Tasks	Semantic Segmentation
Published	2020-03-31
URL	https://arxiv.org/abs/2003.13948v1
PDF	https://arxiv.org/pdf/2003.13948v1.pdf
PWC	https://paperswithcode.com/paper/segmenting-transparent-objects-in-the-wild
Repo
Framework

Efficient Learning of Model Weights via Changing Features During Training


Title	Efficient Learning of Model Weights via Changing Features During Training
Authors	Marcell Beregi-Kovács, Ágnes Baran, András Hajdu
Abstract	In this paper, we propose a machine learning model, which dynamically changes the features during training. Our main motivation is to update the model in a small content during the training process with replacing less descriptive features to new ones from a large pool. The main benefit is coming from the fact that opposite to the common practice we do not start training a new model from the scratch, but can keep the already learned weights. This procedure allows the scan of a large feature pool which together with keeping the complexity of the model leads to an increase of the model accuracy within the same training time. The efficiency of our approach is demonstrated in several classic machine learning scenarios including linear regression and neural network-based training. As a specific analysis towards signal processing, we have successfully tested our approach on the database MNIST for digit classification considering single pixel and pixel-pairs intensities as possible features.
Tasks
Published	2020-02-21
URL	https://arxiv.org/abs/2002.09249v1
PDF	https://arxiv.org/pdf/2002.09249v1.pdf
PWC	https://paperswithcode.com/paper/efficient-learning-of-model-weights-via
Repo
Framework

Network Cooperation with Progressive Disambiguation for Partial Label Learning


Title	Network Cooperation with Progressive Disambiguation for Partial Label Learning
Authors	Yao Yao, Chen Gong, Jiehui Deng, Jian Yang
Abstract	Partial Label Learning (PLL) aims to train a classifier when each training instance is associated with a set of candidate labels, among which only one is correct but is not accessible during the training phase. The common strategy dealing with such ambiguous labeling information is to disambiguate the candidate label sets. Nonetheless, existing methods ignore the disambiguation difficulty of instances and adopt the single-trend training mechanism. The former would lead to the vulnerability of models to the false positive labels and the latter may arouse error accumulation problem. To remedy these two drawbacks, this paper proposes a novel approach termed “Network Cooperation with Progressive Disambiguation” (NCPD) for PLL. Specifically, we devise a progressive disambiguation strategy of which the disambiguation operations are performed on simple instances firstly and then gradually on more complicated ones. Therefore, the negative impacts brought by the false positive labels of complicated instances can be effectively mitigated as the disambiguation ability of the model has been strengthened via learning from the simple instances. Moreover, by employing artificial neural networks as the backbone, we utilize a network cooperation mechanism which trains two networks collaboratively by letting them interact with each other. As two networks have different disambiguation ability, such interaction is beneficial for both networks to reduce their respective disambiguation errors, and thus is much better than the existing algorithms with single-trend training process. Extensive experimental results on various benchmark and practical datasets demonstrate the superiority of our NCPD to other state-of-the-art PLL methods.
Tasks
Published	2020-02-22
URL	https://arxiv.org/abs/2002.11919v1
PDF	https://arxiv.org/pdf/2002.11919v1.pdf
PWC	https://paperswithcode.com/paper/network-cooperation-with-progressive
Repo
Framework

Revisiting SGD with Increasingly Weighted Averaging: Optimization and Generalization Perspectives


Title	Revisiting SGD with Increasingly Weighted Averaging: Optimization and Generalization Perspectives
Authors	Zhishuai Guo, Zixuan Wu, Yan Yan, Xiaoyu Wang, Tianbao Yang
Abstract	Stochastic gradient descent (SGD) has been widely studied in the literature from different angles, and is commonly employed for solving many big data machine learning problems. However, the averaging technique, which combines all iterative solutions into a single solution, is still under-explored. While some increasingly weighted averaging schemes have been considered in the literature, existing works are mostly restricted to strongly convex objective functions and the convergence of optimization error. It remains unclear how these averaging schemes affect the convergence of {\it both optimization error and generalization error} (two equally important components of testing error) for {\bf non-strongly convex objectives, including non-convex problems}. In this paper, we {\it fill the gap} by comprehensively analyzing the increasingly weighted averaging on convex, strongly convex and non-convex objective functions in terms of both optimization error and generalization error. In particular, we analyze a family of increasingly weighted averaging, where the weight for the solution at iteration $t$ is proportional to $t^{\alpha}$ ($\alpha > 0$). We show how $\alpha$ affects the optimization error and the generalization error, and exhibit the trade-off caused by $\alpha$. Experiments have demonstrated this trade-off and the effectiveness of polynomially increased weighted averaging compared with other averaging schemes for a wide range of problems including deep learning.
Tasks
Published	2020-03-09
URL	https://arxiv.org/abs/2003.04339v1
PDF	https://arxiv.org/pdf/2003.04339v1.pdf
PWC	https://paperswithcode.com/paper/revisiting-sgd-with-increasingly-weighted
Repo
Framework